Network Behavioral-Feedback SIR Epidemic Model
Martina Alutto footnoteinfo, Leonardo Cianfanelli, Giacomo Como, Fabio Fagnani · 2025
Read Network Behavioral-Feedback SIR Epidemic Model by Martina Alutto footnoteinfo, Leonardo Cianfanelli, Giacomo Como, Fabio Fagnani (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2507-03852.txt.
Introduction
An epidemic model where spread rates answer back to the outbreak is built by Martina Alutto and co-authors in a 2025 arXiv paper. In their SIR web, groups mix at rates that shift with how sick the web is. That loop stands for steps like distance keeping and mask use. Groups also vary in contact, catch ease, and spread ease.
Stability of rest states is worked out in full. The shape of the safe zone turns on the mix web's build, which guides control design. For one web class, a summed curve always peaks once. That single-peak trait aids later control work.
Catalog record
| CREATOR | Martina Alutto footnoteinfo, Leonardo Cianfanelli, Giacomo Como, Fabio Fagnani |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 39 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2507-03852
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Network Behavioral-Feedback SIR Epidemic Model Martina Alutto footnoteinfo, Leonardo Cianfanelli, Giacomo Como, Fabio Fagnani Abstract We propose a network behavioral-feedback Susceptible-Infected-Recovered (SIR) epidemic model in which the interaction matrix describing the infection rates across subpopulations depends in feedback on the current epidemic state. This model captures both heterogeneities in individuals’ mixing, contact frequency, aptitude to contract and spread the infection, and endogenous behavioral responses such as voluntary social distancing and the adoption of self-protective measures. We study the stability of the equilibria and illustrate through several examples how the shape of the stability region depends on the structure of the interaction matrix, providing insights for the design of effective control strategies. We then analyze the transient behavior of the dynamics, showing that, for a special class of rank-1 interaction matrices, there always exists an aggregate infection curve that exhibits a unimodal behavior, expanding the results on the unimodality of infection curve known in the literature of epidemic models and paving the way for future control applications. keywords Network epidemics model, stability, behavioral-feedback † † thanks: [ † † address: Department of Mathematical Sciences “G.L. Lagrange,” Politecnico di Torino, 10129 Torino, Italy † † address: Department of Automatic Control, Lund University, 22100 Lund, Sweden. † † address: Institute of Electronics, Computer and Telecommunication Engineering, National Research Council of Italy, Politecnico di Torino, 10129 Torino, Italy footnoteinfo]Corresponding author: Martina Alutto. This work was partially supported by the Research Project PRIN 2022 “Extracting Essential Information and Dynamics from Complex Networks” (Grant Agreement number 2022MBC2EZ) funded by the Italian Ministry of University and Research. , , , 1 Introduction Mathematical models play a central role in forecasting the spread of infectious diseases and guiding effective public health interventions. Among these, compartmental models are widely used for the analysis and control of epidemic outbreaks. One of the most studied compartmental models is the Susceptible-Infected-Recovered (SIR) epidemic model, see Kermack & McKendrick 1927 ; Hethcote 2000 ; McKendrick n.d. ; Diekmann & Heesterbeek 2000 . In the SIR epidemic model the population is split into three compartments: susceptible agents who have not yet been infected, infected agents who carry the pathogen and can transmit the disease, and recovered agents, who have healed and gained immunity. The classical SIR model has served as a foundation for many recent control strategies, including those developed during the COVID-19 pandemic, see, e.g., Cianfanelli et al. 2021 ; Miclo et al. 2022 . However, the classical SIR model relies on homogeneity assumptions regarding the pattern of contacts and the transmission rate that are often unrealistic. To address these limitations, network-based epidemic models have been proposed and analyzed in the last years as in Pastor-Satorras et al. 2015 ; Paré et al. 2020 ; Zino & Cao 2021 . In these models, the nodes of the network represent different subpopulations of indistinguishable agents that may differ in several aspects, such as geographical area or age, e.g. in Hethcote 1978 ; Fall et al. 2007 ; Nowzari et al. 2016 ; Mei et al. 2017 ; Ogura & Preciado 2016 . These subpopulations interact through a network, whose links represent the pattern of contacts. The interaction matrix A encodes the infection rate across the subpopulations, taking into account factors such as different susceptibility and infectivity levels of the subpopulations, and different interaction rates among them. Most of the literature on networked epidemic models has focused on the Susceptible-Infected-Susceptible (SIS) epidemic model, where the agents do not acquire immunity after recovering from the disease. The seminal work by Lajmanovich & Yorke 1976 established a key bifurcation result: when the largest eigenvalue of A is below a threshold, the disease-free equilibrium is stable; otherwise, the system admits an endemic equilibrium. Many extensions of the network SIS model have been then studied, including models with higher-order interactions Cisneros-Velarde & Bullo 2021 and competing viruses Anderson & Ye 2023 ; Liu et al. 2019 . The network SIR epidemic model has been the subject of both numerical and theoretical studies, due to its relevance in capturing the spread of epidemics across subpopulations while accounting for recovery-induced immunity. On the numerical side, networked SIR models have been used to evaluate the effectiveness of targeted interventions, for instance to demonstrate that age-specific lockdown policies outperform uniform lockdown policies in reducing mortality and sustaining economic activity, even in simplified settings with two nodes only Acemoglu et al. 2021 . From a theoretical perspective, several works have investigated the network SIR model Mei et al. 2017 ; Ellison 2020 ; Alutto et al. 2025 . Of particular interest is the analysis when the interaction matrix is rank- 1 , which corresponds to assuming that the interaction rate between two subpopulations is proportional to the activity rate of the two subpopulations, without considering homophily Hethcote & Yorke 2014 ; Ellison 2020 ; Alutto et al. 2025 . In particular, Ellison 2020 studied the case of symmetric rank- 1 interaction matrix (corresponding to a uniform size, infectivity level, and susceptibility level of the subpopulations), providing, among other contributions, a characterization of the set of stable equilibria of the model that depends on the heterogeneity of the activity rates of the subpopulations. The analysis of the network SIR model has been largely enriched in Alutto et al. 2025 , where the authors provided a rich set of theoretical results for the case of rank- 1 asymmetric interaction matrix with n nodes. These results include the identification of n invariants of motion, conditions for the stability of the equilibria, a detailed analysis of transient dynamics at the single node level revealing the emergence of multimodal behaviors of the infection curves of single nodes, and the existence of a weighted aggregate infection curve that exhibits a unimodal behavior. Besides the assumption of homogeneous population, another key limitation of the classical SIR model is the assumption of a constant transmission rate, which ignores the fact that typically the agents react to the spread of the epidemic and adapt their behavior accordingly Crosby 2003 ; Lau et al. 2005 . An approach to overcome this limitation is to consider behavioral-feedback epidemic models , which incorporate the human endogenous behavior by letting the infection rate depend in feedback on the current epidemic state Funk et al. 2010 ; Verelst et al. 2016 . The first feedback epidemic model was introduced in Capasso & Serio 1978 , where the authors showed that, if the infection rate is a decreasing function of the fraction of infected agents, the infection curve remains unimodal as in the classical SIR model. Subsequent studies demonstrated analytically that neglecting the human endogenous behavior may lead to mistakes in forecasting the trajectory of the epidemic, as the presence of feedback mechanisms can reduce the peak of the infection curve Baker 2020 ; Gao et al. 2024 ; Franco 2020 . Some models extended the analysis by allowing the transmission rate to depend on both the fraction of susceptible and infected agents Korobeinikov 2006 . In particular, Alutto et al. 2021 proved that the unimodality persists even when the infection rate depends jointly on the fraction of susceptible and infected agents, if the infection rate is non-decreasing in the fraction of susceptible agents. An alternative approach to incorporate